KubeCon Europe is running this week in Amsterdam and NVIDIA used the event to talk up their open-source work around AI and newest open-source contributions.
NVIDIA announced today at KubeCon Europe that they are donating the NVIDIA Dynamic Resource Allocation “DRA” driver to the Cloud Native Computing Foundation. This NVIDIA driver will enjoy “full community ownership” and broader collaboration. The NVIDIA DRA driver for Kubernetes is for configuring and sharing devices like GPUs. The DRA driver allows for sharing and dynamically re-configuring GPUs as well as “ComputeDomains” fr Multi-Node NVLink (MNNVL) usage.
NVIDIA’s intent with this open-source contribution is to “[make] high-performance GPU orchestration seamless and accessible to all.” The NVIDIA DRA driver code is available now on GitHub.
NVIDIA also announced at the event their GPU support for Kata Containers. NVIDIA additionally announced that their KAI Scheduler software has been onboarded as a Cloud Native Computing Foundation Sandbox project.
Today’s NVIDIA announcement also notes how they are “expanding the open-source horizon” with the NVSentinel GPU fault remediation software and AI Cluster Runtime agentic AI framework that they announced last week at GTC. NemoClaw and OpenShell are two of their other open-source projects from GTC that they are also promoting at KubeCon Europe.
